All these systems allow a preoperative or intraoperative planning of the surgery based on patient’s specific information, permitting the simulation of different surgical strategies and to choose the optimal one. In case of preoperative planning, during the intervention, the planned strategy can be displayed in the virtual scenario and integrated with intraoperative information. During the intervention, providing visual and numerical information of the position of surgical tools and patient’s anatomy is updated in real-time, these systems give to the surgeon an accurate visual feedback, guiding him towards the completion of the planned strategy. These systems have the ability to provide more visual information in the surgical field than is available to the naked eye and help in solving the problems of visualization especially in case of minimally invasive surgery. It is well known that less invasive surgery contribute to rather better patient outcomes, but minimization of invasiveness generally results in a lack of surgeon’s perception and dexterity. Showing the structural anatomy of the patient together with surgical tools, these augmented reality navigation systems allow the surgeon to accurately localize anatomical areas and to reach predefined positions watching at the display device.

Several orthopaedic procedures lends themselves well to the use of augmented reality navigation systems. Most common tasks developed by existing navigation systems are: joint replacement, arthroscopic surgery, fracture treatment and spinal surgery.

As proposed in Picard et al. [38], surgical navigation systems can be divided in: (a) systems using preoperative models. They can be either from preoperative three dimensional anatomy models reconstructed from CT/MRI images of the femur and of the tibia of the patient (patient specific) or generic anatomical model of the part in exam (non patient specific). By examining the imaging modalities on which the existing systems are based, one can identify two antithetic trends. The first one is patient oriented and tends to exploit the most advanced devices and techniques to reach excellent quality standards. CT and MRI based systems generally go in this direction and are very powerful, accurate and expensive. The other trend is focused on simplicity and accessibility. Conventional X-ray and fluoroscopy based systems are often less accurate, but affordable for most potential users. The choice of a specific imaging modality is frequently linked with that of a particular registration approach [39]. The image based navigation system has the advantage of precise preoperative planning, but the additional radiation, cost of preoperative CT scans, and additional time for preoperative planning do not improve accuracy [40]. (b) Systems using intraoperative models. They help the surgeon to achieve the planned implant. The knee joint, the data collection probe, and the implant insertion tool are sensorized with securely affixed optical targets and, after the registration and calibration process performed using a particular procedure, they can be localized and tracked in real time (Fig. 25.4). They use intraoperatively acquired medical images to determine the model (image based), and systems using models derived from information determined with direct measurement of the bone surface or limb dynamics (non image based). During surgery, using the localizer, the surgeon collects points on tibial and femoral surfaces, and statistical information about the shape variation of the femur are used to interpolate the data points. The positions of all sensorized objects, including the patient’s anatomy, are tracked in real time by an optoelectronic localizer and reproduced in the virtual scene. Then, through the registration process, the system matches the ideal fitting between acquired points and the generic model. Using a statistical method to build the model has the advantage of requiring less data points to obtain a sufficient interpolation accuracy, reducing intraoperative time required for data acquisition [41]. Finally, to be accepted by surgeons, equipment have to be small unobtrusive, user friendly, safe and compatible to the surgical environment. Adding any equipment in the operating room increases the risk of infection, so any part of the system that comes into contact with the sterile field must be easily sterilized or draped. The digitalization of the bony landmarks is one of the crucial steps in navigation. Problems of reproducibility with intraoperative termination of these landmarks have been described to especially appear with the femoral epicondyles [42, 43].